Colorectal premalignancy is associated with consensus molecular subtypes 1 and 2.

Background: Gene expression-based profiling of colorectal cancer (CRC) can be used to identify four molecularly homogeneous consensus molecular subtype (CMS) groups with unique biologic features. However, its applicability to colorectal premalignant lesions remains unknown. Patients and methods: We assembled the largest transcriptomic premalignancy dataset by integrating different public and proprietary cohorts of adenomatous and serrated polyps from sporadic (N = 311) and hereditary (N = 78) patient populations and carried out a comprehensive analysis of carcinogenesis pathways using the CMS random forest (RF) classifier. Results: Overall, transcriptomic subtyping of sporadic and hereditary polyps revealed CMS2 and CMS1 subgroups as the predominant molecular subtypes in premalignancy. Pathway enrichment analysis showed that adenomatous polyps from sporadic or hereditary cases (including Lynch syndrome) displayed a CMS2-like phenotype with WNT and MYC activation, whereas hyperplastic and serrated polyps with CMS1-like phenotype harbored prominent immune activation. Rare adenomas with CMS4-like phenotype showed significant enrichment for stromal signatures along with transforming growth factor-ß activation. There was a strong association of CMS1-like polyps with serrated pathology, right-sided anatomic location and BRAF mutations. Conclusions: Based on our observations made in premalignancy, we propose a model of pathway activation associated with CMS classification in colorectal carcinogenesis. Specifically, while adenomatous polyps are largely CMS2, most hyperplastic and serrated polyps are CMS1 and may transition into other CMS groups during evolution into carcinomas. Our findings shed light on the transcriptional landscape of premalignant colonic polyps and may help guide the development of future biomarkers or preventive treatments for CRC.

RNA sequencing confirms similarities between PPI-responsive oesophageal eosinophilia and eosinophilic oesophagitis.

BACKGROUND: Although current American guidelines distinguish proton pump inhibitor-responsive oesophageal eosinophilia (PPI-REE) from eosinophilic oesophagitis (EoE), these entities are broadly similar. While two microarray studies showed that they have similar transcriptomes, more extensive RNA sequencing studies have not been done previously. AIM: To determine whether RNA sequencing identifies genetic markers distinguishing PPI-REE from EoE. METHODS: We retrospectively examined 13 PPI-REE and 14 EoE biopsies, matched for tissue eosinophil content, and 14 normal controls. Patients and controls were not PPI-treated at the time of biopsy. We did RNA sequencing on formalin-fixed, paraffin-embedded tissue, with differential expression confirmation by quantitative polymerase chain reaction (PCR). We validated the use of formalin-fixed, paraffin-embedded vs RNAlater-preserved tissue, and compared our formalin-fixed, paraffin-embedded EoE results to a prior EoE study. RESULTS: By RNA sequencing, no genes were differentially expressed between the EoE and PPI-REE groups at the false discovery rate (FDR) ≤0.01 level. Compared to normal controls, 1996 genes were differentially expressed in the PPI-REE group and 1306 genes in the EoE group. By less stringent criteria, only MAPK8IP2 was differentially expressed between PPI-REE and EoE (FDR = 0.029, 2.2-fold less in EoE than in PPI-REE), with similar results by PCR. KCNJ2, which was differentially expressed in a prior study, was similar in the EoE and PPI-REE groups by both RNA sequencing and real-time PCR. CONCLUSION: Eosinophilic oesophagitis and PPI-REE have comparable transcriptomes, confirming that they are part of the same disease continuum.

Evaluation of cytotoxicity, cell proliferation, and genotoxicity induced by p-cresidine in hetero- and nullizygous transgenic p53 mice.

The heterozygous p53 knockout mouse is being used as a short-term alternative model for carcinogenicity screening of chemicals. In most cases, these mice develop tumors within 6 months of exposure to genotoxic carcinogens. The bladder and liver carcinogen, p-cresidine, is recommended as a positive control chemical for these assays. To evaluate early effects of p53 deficiency on bladder and liver histopathology and genotoxicity induced by p-cresidine, we treated 4-week-old heterozygous and nullizygous p53 male mice with p-cresidine by gavage (100, 200, 400, and 800 mg/kg/day) 5 days/week for 7 weeks. Tissue sections were prepared for hematoxylin-eosin staining and immunohistochemistry for PCNA protein or 3'-OH DNA fragments to assess cell proliferation and apoptosis, respectively. Blood and bone marrow were examined for methemoglobin and micronuclei in polychromatic erythrocytes (MN-PCE), respectively. Individual cell necrosis of the bladder transitional epithelium was evident in both p53 heterozygous and nullizygous mice at all doses. In addition, diffuse hyperplasia of the bladder epithelium was observed at 400 and 800 mg/kg in both genotypes. In the liver, both genotypes exhibited similar increases in hepatocyte apoptosis (10-fold increase) and cell proliferation (20-fold increase) at 800 mg/kg/day. Methemoglobin levels were increased 6-fold in both genotypes at 800 mg/kg. Background MN-PCE rates were similar in both genotypes and there were no treatment-related increases. Also, no point mutations were observed in codon 12 of the c-Ha-ras gene from urinary bladder DNA from p-cresidine treated p53 mice. These results suggest that loss of p53 allele(s) in mice does not influence the early markers of carcinogenic activity induced by subchronic treatment with p-cresidine. Increased tumor susceptibility associated with a reduction in p53 dosage may be dependent on neoplastic progression rather than initiation and promotional events elicited by p-cresidine.

Diallyl sulfide enhances azoxymethane-induced preneoplasia in Fischer 344 rat colon.

Azoxymethane (AOM) is an indirect-acting colon carcinogen that produces a high incidence of precancerous lesions, referred to as aberrant crypt foci (ACF), in rats. This study was undertaken to determine whether high dose gavage administration of the cytochrome P-450 2E1 (CYP2E1) inhibitor and chemopreventive agent, diallyl sulfide, would reduce the incidence and severity of ACF formation in the distal colons of AOM-treated Fischer 344 rats. Seven-week-old male rats received 150 or 50 mg/kg diallyl sulfide by gavage 24 and 2 h prior to two weekly i.p. injections of AOM (20 mg/kg). Ten weeks after the last injection of AOM the rats were sacrificed and the colons removed and stained with 0.2% methylene blue. ACF were visualized using stereomicroscopy. Rats pretreated with diallyl sulfide exhibited a significant increase in the number of ACF/cm in the distal colon compared with rats receiving AOM alone. This increase in ACF number was seen in ACF of all sizes. To examine the effects of diallyl sulfide on the initiation stage of AOM-induced carcinogenesis, mutations in the K-ras proto-oncogene were also investigated. ACF and normal appearing colonic mucosa (0.2-0.5 mm3) were microdissected for subsequent PCR-RFLP analysis of a codon 12 (GGT-GGA) activating mutation in the K-ras gene. Greater than 90% of ACF from AOM-treated animals, regardless of diallyl sulfide treatment, exhibited activating K-ras mutations. K-ras mutations were also detected in normal appearing mucosa of AOM-treated animals, although at a lesser frequency (15-35%). These studies demonstrate that diallyl sulfide given in large gavage doses enhances AOM-induced preneoplasia in rats and suggests that diallyl sulfide may alter the disposition of AOM intermediates and/or enhance colonic promotional activity in the rat.

Azoxymethane induces KI-ras activation in the tumor resistant AKR/J mouse colon.

A differential susceptibility phenotype to the organotropic colon carcinogen azoxymethane (AOM) has been described in mice. The following studies were undertaken to test the hypothesis that intraspecific susceptibility can be accounted for by the specific complement of genetic alterations acquired by precancerous colon lesions referred to as aberrant crypt foci (ACF). As an initial approach to this question, mutations in codons 12 and 13 of the Ki-ras proto-oncogene were assessed in ACF, normal-appearing AOM-treated colonic epithelium, and tumors from A/J and SWR/J (susceptible) as well as AKR/J (resistant) mice. Four-week-old male mice were injected intraperitonealy, with AOM once a week for a total of 6 wk and killed 4 and 24 wk after the last injection. DNA was isolated from microdissected tissue, and polymerase chain reaction (PCR)-amplified products of Ki-ras exon 1 (codons 12 and 13) were directly sequenced from microdissected tissues. At 4 wk after AOM exposure, there was no significant difference in the frequency of Ki-ras activation (20-33%) between the three strains. Ki-ras mRNA expression was also evaluated by reverse transcription (RT)-PCR analysis and was comparably reduced (40-50%) in all three strains at the 4 wk time point. However, Ki-ras expression returned to normal by 24 wk after treatment. Finally, to gain further insight into the molecular pathogenesis underlying this experimental tumor model, analysis of the adenomatous polyposis coli (APC) protein within the colonic epithelium was undertaken by using an immunohistochemical approach. Although the APC protein was lost to a varying extent in tumors from A/J and SWR/J mice, the full-length form of the protein was still present in precancerous ACF isolated from each of the three strains, regardless of the degree of dysplasia of the lesion. A further molecular genetic analyses of ACF will be required to gain a more complete understanding of the molecular basis of tumor susceptibility phenotype in this murine model.

V-Ha-ras gene expression in liver and kidney of transgenic Tg.AC mice following chemically induced tissue injury.

The dermal Tg.AC transgenic mouse line is currently being used as a short-term alternative in vivo model for carcinogenicity screening of drugs and environmental chemicals. These mice carry multiple copies of an activated v-Ha-ras oncogene, making them susceptible to promotionally induced tumorigenesis caused by carcinogen exposure or deep skin wounding. Transgene expression is associated with tumor development in these animals. To determine whether tissue injury in organs other than the skin can induce transgene expression, we characterized the pattern of transgene expression in naive animals as well as mice treated by oral gavage with cytotoxic doses of chloroform. Hepatic BrdU labeling was increased 40-fold in females (240 mg/kg/day) and 20-fold in males (140 mg/kg/day) after 4 days of dosing with chloroform. An increase in renal BrdU labeling (7-fold) was observed only in male Tg.AC mice. Although chloroform did not induce v-Ha-ras expression, in either the liver or the kidney, a constitutive amount of transgene message was evident in the kidneys of Tg.AC mice. V-Ha-ras transgene expression also correlated with the expression of GATA-3, a transcription factor that binds the zeta-globin (zeta-globin) promoter of the Tg.AC transgene. These studies suggest that chemically induced tissue injury and regenerative cell proliferation per se are not sufficient for the induction of transgene expression in the liver and kidney of Tg.AC mice. Although organs like the kidney may contain the necessary transcription factors for transgene expression, other factors, yet unidentified, may impede v-Ha-ras-mediated tumorigenesis in these tissues.

Quantitative assessment of azoxymethane-induced aberrant crypt foci in inbred mice.

Heritable differences in tumor susceptibility are observed in mice after repetitive exposures to the organotropic colon carcinogen azoxymethane (AOM). The following study was undertaken to determine whether early morphological alterations within the colonic epithelium correlate with subsequent cancer risk. A/J and SWR/J (susceptible) and AKR/J (resistant) mice were injected once a week with AOM at a dose of 10 mg/kg, i.p., for a total of 6 weeks. Four weeks after the last injection, methylene blue-stained whole-mount colons were examined for the presence of colonic epithelial lesions referred to as aberrant crypt foci (ACF). Putative lesions identified under low magnification were further characterized by H&E staining of corresponding sections. AOM produced a treatment-related increase in ACFs in each of the mouse lines examined. The tumor-susceptible SWR/J and A/J mice developed on average between three- and sixfold more ACFs in the distal colon (32 and 15/cm of colon, respectively) than the resistant AKR/J mice (5/cm colon). The size distribution of ACFs was further analyzed in each of the strains. In SWR/J and A/J, 20-35% of lesions were classified as large ACFs, consisting of 5 or more aberrant crypts per focus. This is in striking contrast to the size distribution of lesions identified in the AKR/J colons, where fewer than 5% of grossly identified lesions were classified as large. In fact, the majority (> 80%) of ACFs in AKR/J mice consisted of only 1-2 aberrant crypts@focus. In addition, there was no evidence of dysplasia in any of the AKR/J lesions examined, whereas the lesions in susceptible mice were dysplastic (adenomas). Our data indicate that tumorigenic response is associated with the extent and multiplicity of ACFs that form within the colonic epithelium at an early time point after carcinogen exposure. These studies further support the use of this morphological biomarker as a short-term endpoint of colon tumorigenesis.

The role of alcohol dehydrogenase in the metabolism of the colon carcinogen methylazoxymethanol.

The aim of this study was to determine whether the cytosolic enzyme alcohol dehydrogenase (ADH) activates methylazoxymethanol (MAM) in the mouse colon and whether differential tumor susceptibility in the mouse is dependent, in part, on strain-related differences in MAM metabolism by ADH. Liver and colon cytosols were isolated from 7-week-old male tumor-susceptible (SWR/J) and -resistant (AKR/J) mice. Minimal reduction of NAD+ was found in colon cytosols from AKR/J mice at the highest concentration (2 mM) of MAM tested. In liver cytosols, only SWR was capable of sustaining NAD+ reduction with MAM, although at very low levels. Despite minimal reactivity with MAM, however, mouse cytosols did effectively reduce NAD+ in the presence of the common ADH subrates ethanol and benzyl alcohol. NAD(+)-coupled oxidation of benzyl alcohol was significantly higher (two- to three-fold, p < 0.05) in mouse colon cytosols compared to activity present within corresponding rat tissues. Incubation of colon and liver cytosols with the ADH-3 inhibitor 4-methylpyrazole markedly (95-100% of controls) reduced ethanol oxidation in both strains. However, 4-methylpyrazole was a less effective inhibitor of benzyl alcohol oxidation in AKR/J colons, suggesting a different ADH isoform complement. An opposite inhibition pattern of benzyl alcohol oxidation was seen in the liver, where 4-methylpyrazole produced a greater inhibition in SWR/J mice. These studies suggest that the metabolism of the proximate mutagen MAM occurs by processes in the mouse that are independent of ADH.

Azoxymethane-induced colon tumors and aberrant crypt foci in mice of different genetic susceptibility.

Azoxymethane (AOM) is an organotropic colon carcinogen that is commonly used to induce colon tumors in rodents. Unlike its parent compound, 1,2-dimethylhydrazine (DMH), a tumor susceptibility phenotype in inbred mice with respect to AOM has not been established. Thus, this study was undertaken to determine whether genetic susceptibility extends to this carcinogen. SWR/J, A/J (both susceptible to DMH carcinogenesis) and AKR/J (resistant) mice were treated with 10 mg/kg AOM i.p. once a week for 8 weeks. Twenty-five weeks after the initial injection, tumor yield was determined. With a single exception, only SWR/J and A/J mice developed tumors, with a distribution that was limited to the distal colon (16.3+/-1.1 and 36.4+/-2.4. respectively). The formation of aberrant crypt foci (ACF), putative preneoplastic lesions, was also assessed in whole-mount colons using Methylene Blue staining. Consistent with tumor multiplicity, the total number of ACF was highest in A/J mice, followed by SWR/J mice. In addition, A/J mice had a significantly greater number of large ACF (five or more crypts per foci) than the other strains. Despite the absence of colon tumors, however, AKR/J mice did develop a significant number of ACF. This finding suggests that ACF in resistant mice are persistent but do not progress to tumors.

Initial levels of azoxymethane-induced DNA methyl adducts are not predictive of tumor susceptibility in inbred mice.

Inbred mice vary in susceptibility to colon carcinogens such as 1,2-dimethylhydrazine (DMH). Differential susceptibility may depend, in part, on formation of promutagenic DNA methyl adducts within target colonic mucosa. The present study was undertaken to evaluate the extent of DNA adduct formation in susceptible (SWR) and resistant (AKR) mice acutely exposed to the colon carcinogen azoxymethane (AOM), a direct metabolite of DMH. In the first experiment, 8-week-old SWR and AKR mice were treated i.p. with 20 mg/kg AOM and sacrificed 6 h later. DNA was isolated from distal colon and liver, and O6-methylguanine (O6-MeGua) adduct levels were assessed by immunoslot blot (ISB) analysis, using a monospecific antibody raised against O6-methyldeoxyguanosine. HPLC-fluorescence detection was also used to quantitate 06-MeGua and 7-methylguanine (7-MeGua), and to generate standard curves. At 6 h, both O6-MeGua and 7-MeGua were significantly higher (2- to 3-fold, p < 0.05) in AKR colon, while an opposite pattern was found in liver. In Experiment 2, mice were injected with AOM (20 mg/kg) and euthanized 12 and 48 h later. At 12 h, O6-MeGua levels were higher in colons (1.4-fold) of SWR mice. Forty-eight hours after treatment, however, adduct levels in colon were markedly (5-fold) reduced in SWR but were unchanged from 12 h in AKR. To further compare activation of AOM in both strains, colon microsomes were incubated with AOM and calf thymus DNA. Comparable levels of O6-MeGua were detected by ISB, demonstrating equivalent metabolic capacity in both SWR and AKR mice. These studies suggest that differential susceptibility to AOM-induced colon carcinogenesis is not based on initial target tissue DNA alkylation and unlikely to depend on differential metabolic capacity.

A comparative study of hepatic and colonic metabolic enzymes in inbred mouse lines before and after treatment with the colon carcinogen, 1,2-dimethylhydrazine.

1,2-Dimethylhydrazine (DMH) is an organotropic colon carcinogen that undergoes metabolic activation to DNA-reactive metabolites. Twenty hours after parenteral treatment of AKR/J (colon tumor resistant) and SWR/J (susceptible) mice with DMH.2HCl (70 mg/kg), functional levels of Cyp1a1 and Cyp2e1 were examined by measuring O-deethylation of ethoxyresorufin (EROD) and hydroxylation of p-nitrophenol, respectively. In control animals, SWR/J mice exhibited higher hepatic EROD activity (1.4-fold) when compared with AKR/J mice. In carcinogen-treated animals, EROD activity was decreased 20-30% in both mouse lines. Hepatic p-nitrophenol hydroxylase activity, similar in control animals of both strains, was reduced comparably (45-50% of control) after DMH administration. In liver, a decrease in immunoreactive Cyp2e1 protein paralleled the decline in enzyme activity, whereas in the colon, no significant treatment-related differences were detected in either strain. In liver and colon cytosols, alcohol dehydrogenase activity was not significantly different in either mouse line, both in control and DMH-treated animals. Glutathione levels were elevated (1.7-fold) in livers of AKR/J mice after DMH administration. Total glutathione-S-transferase (GST) activity was significantly increased (1.8-fold) in the colons of SWR/J mice and in the livers (1.4-fold) of AKR/J mice. Furthermore, the GST isoform, GST-Yp, was reduced 40% in the SWR/J colon. These data demonstrate the importance of metabolic capacity as a factor in conferring differential tumor susceptibility in a murine cancer model to the indirect-acting colon carcinogen, DMH.